In FreeBSD you can share ZFS directly into a jail, allowing the "root" of the jail to manage their own volumes, snapshots, etc, but the host can still maintain restrictions on the jail.
Another fun fact about jails. The host can configure how many jails can be in a jail. Because jails act like a virtualized system, you can just keep chaining jails under jails, each jail can have its own root user, and with ZFS, each jail can manage its own volumes. There is still some work with ZFS resources management that needs to be done to keep jails from DOS'n the host, but you can see how flexible this system is.
As long as current bandwidth is less than 80% of capacity, latency should not be an issue. Bitch at your ISP for not having enough capacity, not at Netflix users for attempting to use what they paid for.
If everyone suddenly [insert impossible thing] at the same time, it would be too expensive to handle that. We've got a genius here, telling us that if impossible things happened, like everyone using the Internet at the same time and attempting to use full speed, that an ISP can't handle that. NO ISP anywhere can handle that, yet you can purchase "dedicated" bandwidth all the same and get crazy awesome performance.
ISPs can supply more bandwidth than you can use while still not providing everyone the ability to use 100% all the time. It's like saying, "they say it's an all you can eat buffet, but if I went there and ate $10,000 of rice in 1 hour, they'd stop serving me".. WTF is wrong with you?! Stop using impossible things are you reason to support crappy internet service.
The Internet has no issues with current loads, bad ISPs have issues with current loads. I pay $90/m for a 100/100 dedicated connection with no cap and I've called my ISP over 10ms ping increases and they've transferred me to an engineer to figure out the issue. Quality bandwidth is cheap. I have quality graphs where my 24x7 1 second ping hadn't lost a single packet in over a month. My average packet-loss is under 5 packets per week, my average ping to my ISP is under 0.2ms, I seed BitTorrent 24x7. My biggest issue is that YouTube's Chicago PoP can burst at least 1Gb/s at me, which forced me to increase my buffer size because I had my buffers configured for 100Mb/s and my ISP's Cisco core router's rate limiter doesn't react fast enough to micro bursts.
To top it off, my ISP uses an AQM, so I can maintain a solid sub 1ms ping with virtually no jitter(under 1ms) or packet-loss while maxing out my connection in both directions. It's not QoS, it's not traffic shaping, it's just one of those nifty new Active Queue Managers that use fair queuing. I recently gave it a test, I uncapped my Torrent client, disabled traffic shaping and let it attempt to download as fast as it could at 9pm. I had a minimum of 98.6Mb/s(1min avg) and a maximum of 99.8Mb/s(1min avg) during a 20 minute window at the peak hours of the night, while getting less than 0.04% packet-loss to a Chicago internet server and maintaining under 1ms of jitter. 0.04% loss is high for me, so I traffic shape to keep it less than 0.00%.
Again, $90/month, not an intro price, and not bundled. My ISP is almost as old as AT&T, they're not going anywhere. In fact, several years back they had to push back my fiber install date a few months because of "unexpectedly high demand". I've never seen so much spam from Charter trying to keep their customer base.
QoS is pointless when you have enough bandwidth. Your VoIP packets are no more important than my VPN packets. Maybe I am using VoIP and doing a file transfer at the same time, through a VPN tunnel, the ISP can't know that. Everyone should just get their fair share and should manage their own bandwidth usage.
At home I use traffic shaping and I can maintain 98% line saturation while still getting less than 0.2ms of jitter.
The biggest complaint of congestion is that it is associated with latency. We have the tech to keep latency low during high saturation, and not even use QoS or even traffic shaping. Look up fq_codel, it's like magic. It's simple, it's "fair"(mostly), it's stateless, it's protocol agnostic, and it's turn-key easy to setup. More ISPs need to use it.
The average person can't manage their own healthcare. The do so at the detriment of all others. It costs more to not have people on insurance. Ideally, we'd have a national insurance that covered everyone via taxes, but we have a hybrid system. The biggest issues with ideals is that they're not always practical.
No matter what, we benefit by having more people with quality insurance. Just with the number of reduced trips to the ER for preventable issues, it will pay itself back in spades. Trips to the ER are magnitudes more expensive than going to the doctor for a scheduled appointment.
the other big issue is that people are worth more than they get paid. A person making $40k/year may be worth over $100k/year to the economy as a whole. If people got paid what their actual value was, there wouldn't be an issue in the first place.
An aggressive alien race that works well enough together to be able to travel light years is unlikely. The bigger issue would not them being aggressive, but too logical. They may see us as a threat because we are prone to outbursts. If we stifle our outbursts, we'd be less of a threat.
A local host doesn't always mean "faster" for either bandwidth or latency. I've seen situations where others had lower pings to me with a different ISP than their neighbors on the same ISP, because of poor routing, and that's not including that they have highly asymmetrical down:up and I have symmetrical.
A simple change would be to make clients favor lower latency clients. Something like floor(log2(latency)), lower is better.
where we think the OS will solve a problem it cannot solve in the general case, is brittle and fails frequently
If your OS is so bad that you don't trust it's entropy, then STOP using that OS. If you can't, then sucks to be you, it's a crap OS and the programmers should die in a fire.
I'm sure they can afford to place a diode on the SSD to keep power from back-flowing, and a sensor the monitor the line voltage on the PSU side of the diode. If the voltage drops too low, the SSD does some quick clean up and sleeps itself.
You're highly unlikely to get corrupted blocks written and just likely to have "missing" data that never got written at all. If your blocks got corrupted, there's a good chance the entire SSD would be bricked or need to be "reset", losing all data.
At least good PSUs don't give you "power flickage". Since they're switching power supplies, once there is not enough power to run the electronics, it dies all at once instead of sagging. I also tend to purchase high end PSUs and they can typically handle the PSU at 100% load for 10ms without power from mains and still maintaining stable clean power. Once the voltage in the caps drops too low, the "switching" part of the PSU stops supplying power.
The proposed net neutrality only applies to the last mile. Don't mess with retail customers, and do whatever you want to commercial customers, as long as it doesn't "harm" retail customers.
Tyndallize requires a device to be pressure cooked at 115c for 15 minute sessions, 3 days in a row. Making devices withstand pressure and heat would increase costs quite a bit, plus the cost of having each device out of service for 3 days between uses. Sounds decent enough for simple tools that are mostly metal, but more complex devices, like endoscopes may have issues.
I forgot to mention, most of the matter in the accretion disk never even makes it into the black hole, but instead gets shot out the jets because of the strong magnetic field.
The accretion disk slows down matter consumption. It's entirely possible that a black hole may able to absorb dark matter more quickly than regular matter because of the lack of interaction with electromagnetism.
Slashdot Mirror
That was when AMD's x64 architecture was designed by an ex-senior Dec Alpha architect.
In FreeBSD you can share ZFS directly into a jail, allowing the "root" of the jail to manage their own volumes, snapshots, etc, but the host can still maintain restrictions on the jail.
Another fun fact about jails. The host can configure how many jails can be in a jail. Because jails act like a virtualized system, you can just keep chaining jails under jails, each jail can have its own root user, and with ZFS, each jail can manage its own volumes. There is still some work with ZFS resources management that needs to be done to keep jails from DOS'n the host, but you can see how flexible this system is.
As long as current bandwidth is less than 80% of capacity, latency should not be an issue. Bitch at your ISP for not having enough capacity, not at Netflix users for attempting to use what they paid for.
If everyone suddenly [insert impossible thing] at the same time, it would be too expensive to handle that. We've got a genius here, telling us that if impossible things happened, like everyone using the Internet at the same time and attempting to use full speed, that an ISP can't handle that. NO ISP anywhere can handle that, yet you can purchase "dedicated" bandwidth all the same and get crazy awesome performance.
ISPs can supply more bandwidth than you can use while still not providing everyone the ability to use 100% all the time. It's like saying, "they say it's an all you can eat buffet, but if I went there and ate $10,000 of rice in 1 hour, they'd stop serving me".. WTF is wrong with you?! Stop using impossible things are you reason to support crappy internet service.
The Internet has no issues with current loads, bad ISPs have issues with current loads. I pay $90/m for a 100/100 dedicated connection with no cap and I've called my ISP over 10ms ping increases and they've transferred me to an engineer to figure out the issue. Quality bandwidth is cheap. I have quality graphs where my 24x7 1 second ping hadn't lost a single packet in over a month. My average packet-loss is under 5 packets per week, my average ping to my ISP is under 0.2ms, I seed BitTorrent 24x7. My biggest issue is that YouTube's Chicago PoP can burst at least 1Gb/s at me, which forced me to increase my buffer size because I had my buffers configured for 100Mb/s and my ISP's Cisco core router's rate limiter doesn't react fast enough to micro bursts.
To top it off, my ISP uses an AQM, so I can maintain a solid sub 1ms ping with virtually no jitter(under 1ms) or packet-loss while maxing out my connection in both directions. It's not QoS, it's not traffic shaping, it's just one of those nifty new Active Queue Managers that use fair queuing. I recently gave it a test, I uncapped my Torrent client, disabled traffic shaping and let it attempt to download as fast as it could at 9pm. I had a minimum of 98.6Mb/s(1min avg) and a maximum of 99.8Mb/s(1min avg) during a 20 minute window at the peak hours of the night, while getting less than 0.04% packet-loss to a Chicago internet server and maintaining under 1ms of jitter. 0.04% loss is high for me, so I traffic shape to keep it less than 0.00%.
Again, $90/month, not an intro price, and not bundled. My ISP is almost as old as AT&T, they're not going anywhere. In fact, several years back they had to push back my fiber install date a few months because of "unexpectedly high demand". I've never seen so much spam from Charter trying to keep their customer base.
QoS is pointless when you have enough bandwidth. Your VoIP packets are no more important than my VPN packets. Maybe I am using VoIP and doing a file transfer at the same time, through a VPN tunnel, the ISP can't know that. Everyone should just get their fair share and should manage their own bandwidth usage.
At home I use traffic shaping and I can maintain 98% line saturation while still getting less than 0.2ms of jitter.
The biggest complaint of congestion is that it is associated with latency. We have the tech to keep latency low during high saturation, and not even use QoS or even traffic shaping. Look up fq_codel, it's like magic. It's simple, it's "fair"(mostly), it's stateless, it's protocol agnostic, and it's turn-key easy to setup. More ISPs need to use it.
The average person can't manage their own healthcare. The do so at the detriment of all others. It costs more to not have people on insurance. Ideally, we'd have a national insurance that covered everyone via taxes, but we have a hybrid system. The biggest issues with ideals is that they're not always practical.
No matter what, we benefit by having more people with quality insurance. Just with the number of reduced trips to the ER for preventable issues, it will pay itself back in spades. Trips to the ER are magnitudes more expensive than going to the doctor for a scheduled appointment.
the other big issue is that people are worth more than they get paid. A person making $40k/year may be worth over $100k/year to the economy as a whole. If people got paid what their actual value was, there wouldn't be an issue in the first place.
Aggression to some extent is a requirement when dealing with limit resources. Because of this, some amount of aggression is naturally selected for.
I wonder how a psychologist would compare competitive and aggressive. Is one a subset of the other?
An aggressive alien race that works well enough together to be able to travel light years is unlikely. The bigger issue would not them being aggressive, but too logical. They may see us as a threat because we are prone to outbursts. If we stifle our outbursts, we'd be less of a threat.
It does lead to all kinds of health complications. Risks increase dramatically across the board.
A local host doesn't always mean "faster" for either bandwidth or latency. I've seen situations where others had lower pings to me with a different ISP than their neighbors on the same ISP, because of poor routing, and that's not including that they have highly asymmetrical down:up and I have symmetrical.
A simple change would be to make clients favor lower latency clients. Something like floor(log2(latency)), lower is better.
No one trusts hardware anymore. It can be used to mix with the entropy pool, but it should never be used directly.
where we think the OS will solve a problem it cannot solve in the general case, is brittle and fails frequently
If your OS is so bad that you don't trust it's entropy, then STOP using that OS. If you can't, then sucks to be you, it's a crap OS and the programmers should die in a fire.
I'm sure they can afford to place a diode on the SSD to keep power from back-flowing, and a sensor the monitor the line voltage on the PSU side of the diode. If the voltage drops too low, the SSD does some quick clean up and sleeps itself.
You're highly unlikely to get corrupted blocks written and just likely to have "missing" data that never got written at all. If your blocks got corrupted, there's a good chance the entire SSD would be bricked or need to be "reset", losing all data.
At least good PSUs don't give you "power flickage". Since they're switching power supplies, once there is not enough power to run the electronics, it dies all at once instead of sagging. I also tend to purchase high end PSUs and they can typically handle the PSU at 100% load for 10ms without power from mains and still maintaining stable clean power. Once the voltage in the caps drops too low, the "switching" part of the PSU stops supplying power.
Net neutrality also states you cannot block or degrade legal traffic.
Why not just shape your own traffic? You get more control and they can't abuse it.
The proposed net neutrality only applies to the last mile. Don't mess with retail customers, and do whatever you want to commercial customers, as long as it doesn't "harm" retail customers.
BitTorrent is something like 10%-20% of peak traffic, I'm not sure why they're so worried about it.
Tyndallize requires a device to be pressure cooked at 115c for 15 minute sessions, 3 days in a row. Making devices withstand pressure and heat would increase costs quite a bit, plus the cost of having each device out of service for 3 days between uses. Sounds decent enough for simple tools that are mostly metal, but more complex devices, like endoscopes may have issues.
All points in space have volume. The smallest anything can be is a cubic plank unit, which still has volume.
I forgot to mention, most of the matter in the accretion disk never even makes it into the black hole, but instead gets shot out the jets because of the strong magnetic field.
The accretion disk slows down matter consumption. It's entirely possible that a black hole may able to absorb dark matter more quickly than regular matter because of the lack of interaction with electromagnetism.